Prospective assessment of accuracy of endoanal MR imaging and endosonography in patients with fecal incontinence

OBJECTIVE. Endoanal MR imaging was prospectively compared with anal endoson ography to determine any superiority in the characterization of sphincter m orphology in fecal incontinence. SUBJECTS AND METHODS. Fifty-two consecutive patients with fecal incontinenc e were examined with anal endosonography and endoanal MR imaging after a de tailed bowel history, clinical examination, and complete anorectal physiolo gic testing. External and internal anal sphincter integrity was noted on bo th endosonograms and MR images by two radiologists in consensus, who read i ndividual scans in a random order to avoid recall bias, Imaging findings we re subsequently compared, and arbitration of any disagreement between endos onography and MR imaging was made in consensus by a surgeon and a gastroent erologist who also had access to the patient's history, clinical examinatio n, and anorectal physiologic testing results. RESULTS. Complete agreement was found between anal endosonographic and MR i maging interpretations in 32 patients (62%): 10 with combined external and internal sphincter injuries, two with isolated internal sphincter injury, a nd 20 with intact sphincters. Of 20 patients in whom results of the scans w ere disparate, incorrect interpretation was found on endosonography in six patients, on MR imaging in 15. Overall, one error relating to the internal sphincter was made on endosonography versus 12 on MR imaging (p = 0.002), a nd five errors relating to the external sphincter were made on endosonograp hy versus six on MR imaging (p = 1.0). CONCLUSION. This study suggests that endoanal sonography and endoanal MR im aging are equivalent in diagnosing external anal sphincter injury, but MR i maging is inferior in diagnosing internal anal sphincter injury.